*BSD News Article 3732


Return to BSD News archive

Path: sserve!manuel!munnari.oz.au!mips!mips!sdd.hp.com!wupost!uunet!stanford.edu!rutgers!igor.rutgers.edu!athos.rutgers.edu!hedrick
From: hedrick@athos.rutgers.edu (Charles Hedrick)
Newsgroups: comp.unix.bsd
Subject: Re: Restrictions on 'free' UNIX / 386BSD (Linux licensing)
Message-ID: <Aug.16.20.35.57.1992.15736@athos.rutgers.edu>
Date: 17 Aug 92 00:35:58 GMT
References: <l8n8qcINN2c5@neuro.usc.edu> <PHR.92Aug15151100@soda.berkeley.edu> <63DILTJ@taronga.com> <1992Aug16.210036.17095@fcom.cc.utah.edu>
Organization: Rutgers Univ., New Brunswick, N.J.
Lines: 45

terry@cs.weber.edu (A Wizard of Earth C) writes:

>	Or, put another way:  It's hard to contribute code to Linux and
>Berkeley (or anything else) at the same time.

This seems to be wrong.  It's attributing a far greater degree
of coordination to Linux than actually exists.

Linus himself sort of uses the GPL.  The kernel code simply says (C)
1991 Linus Torvalds.  He has asserted separately that it may be
redistributed under the terms of the GPL.  But I believe other terms
could be negotiated as well.  Linus has no objections to commercial
distributions of Linux.  He is certainly not the extremist that rms
is.  There has never been a requirement that all Linux code must be
covered by the GPL.  Pieces of code from authors other than Linus
carry whatever terms the authors choose, although I think it's assumed
by everyone that code will be freely redistriutable.  There is
certainly Berkeley code used with Linux, though not in the kernel.
(This may change with TCP/IP support in the kernel -- I don't know
what they are doing, but I have assumed it would be based on some
version of Berkeley's TCP/IP.)

What is true is that code contributed to the Linux kernel will end up
with Linus' copyright being added to it, and rewritten to fit his
coding style and his views of the way the kernel should work.  That
does not affect your ability to continue redistributing the original.
It also assumes that the code is integrated by Linus.  It's perfectly
acceptable to supply additional code that users will have to install
for themselves.

There are no constraints on things outside the kernel.  Outside the
kernel Linux is a typical network hackers' product: There's no central
authority, and no official distributions.  There are people who put
together root disks roughly in sync with the kernel, but those people
are not under Linus' control, and as far as I can tell they take the
best software they can find without any particular requirements on
copyright aside from being redistributable.  Somebody else coordinates
the libraries and GCC.  Since they are based on glibc and gcc,
obviously most of this is covered by the GPL.  A few portions of the
library are taken from BSD.  Those particular modules continue to be
governed by the BSD copyright.  The nearest equivalent to the full
386BSD release is something done by the Manchester Computing
Consortium.  It's a packaged release which seems to have a good
reputation.  But they have no official status, and there are a couple
of other people trying to do their own releases.